A computer vision based human-robot interface

  • Authors:
  • José M. Buenaposada;Luis Baumela

  • Affiliations:
  • Universidad Politécnica de Madrid, Departamento de Inteligencia Artificial, Campus de Montegancedo s/n, 28660 Madrid, Spain;Universidad Politécnica de Madrid, Departamento de Inteligencia Artificial, Campus de Montegancedo s/n, 28660 Madrid, Spain

  • Venue:
  • Autonomous robotic systems
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

This chapter focuses on the real-time location and tracking of human faces in video sequences. The tracking is based on the cooperation of two low-level trackers based on colour and template information. The colour-based tracker is fast and robust, but it can only compute the 2D location of the face on the image. The template-based tracker, although it is more sensitive to environmental variations and more time consuming, it can compute the position and orientation of a human face in 3D space. As a result of the co-ordination of these two trackers, it emerges a robust real-time tracker that accurately computes face position and orientation in varying environmental conditions.